diff --git a/tests/v1/engine/test_core_engine_actor_manager.py b/tests/v1/engine/test_core_engine_actor_manager.py index a986bc07a3..a7a25ef0d6 100644 --- a/tests/v1/engine/test_core_engine_actor_manager.py +++ b/tests/v1/engine/test_core_engine_actor_manager.py @@ -308,7 +308,6 @@ def test_ray_dp_addresses_resolved_before_actor_creation( executor_class=_DummyExecutor, log_stats=False, addresses=addresses, - num_api_servers=2, ) as ( engine_manager, _coordinator, diff --git a/vllm/entrypoints/cli/serve.py b/vllm/entrypoints/cli/serve.py index 08cb79f208..e8f72ab230 100644 --- a/vllm/entrypoints/cli/serve.py +++ b/vllm/entrypoints/cli/serve.py @@ -326,9 +326,12 @@ def run_multi_api_server(args: argparse.Namespace): defer_api_server_ports=not (rust_frontend_path or is_ray_dp), ) - with launch_core_engines( - vllm_config, executor_class, log_stats, addresses, num_api_servers - ) as (local_engine_manager, coordinator, addresses, tensor_queue): + with launch_core_engines(vllm_config, executor_class, log_stats, addresses) as ( + local_engine_manager, + coordinator, + addresses, + tensor_queue, + ): stats_update_address = ( coordinator.get_stats_publish_address() if coordinator else None ) diff --git a/vllm/v1/engine/utils.py b/vllm/v1/engine/utils.py index 9b3bea0db9..af62ed3192 100644 --- a/vllm/v1/engine/utils.py +++ b/vllm/v1/engine/utils.py @@ -1056,7 +1056,6 @@ def launch_core_engines( executor_class: type[Executor], log_stats: bool, addresses: EngineZmqAddresses, - num_api_servers: int = 1, ) -> Iterator[ tuple[ CoreEngineProcManager | CoreEngineActorManager | None,